Zomato Rating Prediction and Restaurant Recommender

Business Problem

With the onset of the pandemic in 2020, The way India eats has changed. The penetration of online food services in India is set to double by 2025. Restaurants will have to soon find a way to control their operating expenses such as rentals, electricity, and manpower costs. Cloud kitchen or delivery only outlets are a viable option for restaurants to explore now. This comes with its challenges; people now are ordering food based on reviews and ratings. Various factors affect the rating of a restaurant apart from quality and taste like the average cost of food, cuisines offered, restaurant type, location etc. new restaurants that have just started up might lack the necessary data to forecast how their restaurant is going to perform and they do trial and error to figure out what works to get good reviews, this takes up a lot of time and resources which many small restaurants can’t afford and thus they might not be able to compete with well-known established restaurants and this might result in restaurants failure. Also, new restaurants face a lot of questions starting from deciding on the location to choosing the type of cuisines to offer. This project helps solve that issue by analyzing the data from Zomato to predict a rating based on certain features and understand the taste of the Bengalureans thus helping restaurants succeed by making the right choices.

Variables & Descriptions:

---

Variables	Description
url:	contains the url of the restaurant in the zomato website
address:	contains the address of the restaurant in Bangalore
name:	contains the name of the restaurant in Bangalore
online_order:	whether online ordering is available in the restaurant or not
book_table:	table book option is available in the restaurant or not
rate:	contains the overall rating of the restaurant out of 5
votes:	contains total number of rating for the restaurant
phone:	contains the phone number of the restaurant
location:	contains the neighbourhood in which the restaurant is located
rest_type:	restaurant type
dish_liked:	dishes liked by people in the restaurant
cuisines:	food styles offered
approx_cost(for two people):	contains the approximate cost of meal for two people
reviews_list:	list of tuples containing reviews for the restaurant, each tuple
menu_item:	contains list of menus available in the restaurant
listed_in(city):	contains the neighborhood in which the restaurant is listed

Importing Libraries

import numpy as np 
import pandas as pd 
import matplotlib.pyplot as plt
%matplotlib inline
import seaborn as sns
import plotly.express as px
import plotly.graph_objs as go
from plotly.offline import iplot
from plotly import tools
import plotly.io as pio
pio.renderers.default = "notebook+pdf" 
from warnings import filterwarnings
filterwarnings('ignore')

Reading Data

df = pd.read_csv(r'D:\Data_science\Projects\Imarticus\Capstone\Dataset\zomato.csv\zomato.csv')

#visualizing the first 5 rows of the dataset to get a overview of the data
df.head()

	url	address	name	online_order	book_table	rate	votes	phone	location	rest_type	dish_liked	cuisines	approx_cost(for two people)	reviews_list	menu_item	listed_in(type)	listed_in(city)
0	https://www.zomato.com/bangalore/jalsa-banasha...	942, 21st Main Road, 2nd Stage, Banashankari, ...	Jalsa	Yes	Yes	4.1/5	775	080 42297555\r\n+91 9743772233	Banashankari	Casual Dining	Pasta, Lunch Buffet, Masala Papad, Paneer Laja...	North Indian, Mughlai, Chinese	800	[('Rated 4.0', 'RATED\n A beautiful place to ...	[]	Buffet	Banashankari
1	https://www.zomato.com/bangalore/spice-elephan...	2nd Floor, 80 Feet Road, Near Big Bazaar, 6th ...	Spice Elephant	Yes	No	4.1/5	787	080 41714161	Banashankari	Casual Dining	Momos, Lunch Buffet, Chocolate Nirvana, Thai G...	Chinese, North Indian, Thai	800	[('Rated 4.0', 'RATED\n Had been here for din...	[]	Buffet	Banashankari
2	https://www.zomato.com/SanchurroBangalore?cont...	1112, Next to KIMS Medical College, 17th Cross...	San Churro Cafe	Yes	No	3.8/5	918	+91 9663487993	Banashankari	Cafe, Casual Dining	Churros, Cannelloni, Minestrone Soup, Hot Choc...	Cafe, Mexican, Italian	800	[('Rated 3.0', "RATED\n Ambience is not that ...	[]	Buffet	Banashankari
3	https://www.zomato.com/bangalore/addhuri-udupi...	1st Floor, Annakuteera, 3rd Stage, Banashankar...	Addhuri Udupi Bhojana	No	No	3.7/5	88	+91 9620009302	Banashankari	Quick Bites	Masala Dosa	South Indian, North Indian	300	[('Rated 4.0', "RATED\n Great food and proper...	[]	Buffet	Banashankari
4	https://www.zomato.com/bangalore/grand-village...	10, 3rd Floor, Lakshmi Associates, Gandhi Baza...	Grand Village	No	No	3.8/5	166	+91 8026612447\r\n+91 9901210005	Basavanagudi	Casual Dining	Panipuri, Gol Gappe	North Indian, Rajasthani	600	[('Rated 4.0', 'RATED\n Very good restaurant ...	[]	Buffet	Banashankari

#visualizing the last 5 rows of the dataset to get a overview of the data
df.tail()

	url	address	name	online_order	book_table	rate	votes	phone	location	rest_type	dish_liked	cuisines	approx_cost(for two people)	reviews_list	menu_item	listed_in(type)	listed_in(city)
51712	https://www.zomato.com/bangalore/best-brews-fo...	Four Points by Sheraton Bengaluru, 43/3, White...	Best Brews - Four Points by Sheraton Bengaluru...	No	No	3.6 /5	27	080 40301477	Whitefield	Bar	NaN	Continental	1,500	[('Rated 5.0', "RATED\n Food and service are ...	[]	Pubs and bars	Whitefield
51713	https://www.zomato.com/bangalore/vinod-bar-and...	Number 10, Garudachar Palya, Mahadevapura, Whi...	Vinod Bar And Restaurant	No	No	NaN	0	+91 8197675843	Whitefield	Bar	NaN	Finger Food	600	[]	[]	Pubs and bars	Whitefield
51714	https://www.zomato.com/bangalore/plunge-sherat...	Sheraton Grand Bengaluru Whitefield Hotel & Co...	Plunge - Sheraton Grand Bengaluru Whitefield H...	No	No	NaN	0	NaN	Whitefield	Bar	NaN	Finger Food	2,000	[]	[]	Pubs and bars	Whitefield
51715	https://www.zomato.com/bangalore/chime-sherato...	Sheraton Grand Bengaluru Whitefield Hotel & Co...	Chime - Sheraton Grand Bengaluru Whitefield Ho...	No	Yes	4.3 /5	236	080 49652769	ITPL Main Road, Whitefield	Bar	Cocktails, Pizza, Buttermilk	Finger Food	2,500	[('Rated 4.0', 'RATED\n Nice and friendly pla...	[]	Pubs and bars	Whitefield
51716	https://www.zomato.com/bangalore/the-nest-the-...	ITPL Main Road, KIADB Export Promotion Industr...	The Nest - The Den Bengaluru	No	No	3.4 /5	13	+91 8071117272	ITPL Main Road, Whitefield	Bar, Casual Dining	NaN	Finger Food, North Indian, Continental	1,500	[('Rated 5.0', 'RATED\n Great ambience , look...	[]	Pubs and bars	Whitefield

#understanding the shape of the dataset
print("The dimension of the dataset: ", df.shape)
print("Number of features: ", df.shape[1])
print("Number of rows: ", df.shape[0])

The dimension of the dataset:  (51717, 17)
Number of features:  17
Number of rows:  51717

#renaming the columns for better readabilty
df = df.rename(columns={'approx_cost(for two people)':'approx_cost','listed_in(type)':'type','listed_in(city)':'listed_city'})

# checking the names of all the features
df.columns

Index(['url', 'address', 'name', 'online_order', 'book_table', 'rate', 'votes',
       'phone', 'location', 'rest_type', 'dish_liked', 'cuisines',
       'approx_cost', 'reviews_list', 'menu_item', 'type', 'listed_city'],
      dtype='object')

# Checking the datatype and presence of null values
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 51717 entries, 0 to 51716
Data columns (total 17 columns):
 #   Column        Non-Null Count  Dtype 
---  ------        --------------  ----- 
 0   url           51717 non-null  object
 1   address       51717 non-null  object
 2   name          51717 non-null  object
 3   online_order  51717 non-null  object
 4   book_table    51717 non-null  object
 5   rate          43942 non-null  object
 6   votes         51717 non-null  int64 
 7   phone         50509 non-null  object
 8   location      51696 non-null  object
 9   rest_type     51490 non-null  object
 10  dish_liked    23639 non-null  object
 11  cuisines      51672 non-null  object
 12  approx_cost   51371 non-null  object
 13  reviews_list  51717 non-null  object
 14  menu_item     51717 non-null  object
 15  type          51717 non-null  object
 16  listed_city   51717 non-null  object
dtypes: int64(1), object(16)
memory usage: 6.7+ MB

It can be seen from the above output that some features contain null values we need to treat those values

# Checking count and percentage of null values of each feature

total = df.isnull().sum().sort_values(ascending=False)
percent = ((df.isnull().sum()/df.isnull().count()*100)).sort_values(ascending=False)
missing_data = pd.concat([total, percent], axis=1, keys=['Total', 'Percent'])
missing_data.head(17)

	Total	Percent
dish_liked	28078	54.291626
rate	7775	15.033741
phone	1208	2.335789
approx_cost	346	0.669026
rest_type	227	0.438927
cuisines	45	0.087012
location	21	0.040606
type	0	0.000000
menu_item	0	0.000000
reviews_list	0	0.000000
url	0	0.000000
address	0	0.000000
votes	0	0.000000
book_table	0	0.000000
online_order	0	0.000000
name	0	0.000000
listed_city	0	0.000000

# visualising null values

sns.heatmap(df.isnull())
plt.show()

'dish_liked' has about 54% of missing values if we try to handle this missing value we will introduce bias. Hence we are dropping the column

#dropping the 'dish_liked' column
df.drop('dish_liked', axis = 1, inplace = True)

df.shape # we have succesfully dropped the 'dish_liked' column

(51717, 16)

url and phone does not add any value to the rating hence we will drop those columns as well

df.drop(['url', 'phone'], axis = 1, inplace = True)

df.head()

	address	name	online_order	book_table	rate	votes	location	rest_type	cuisines	approx_cost	reviews_list	menu_item	type	listed_city
0	942, 21st Main Road, 2nd Stage, Banashankari, ...	Jalsa	Yes	Yes	4.1/5	775	Banashankari	Casual Dining	North Indian, Mughlai, Chinese	800	[('Rated 4.0', 'RATED\n A beautiful place to ...	[]	Buffet	Banashankari
1	2nd Floor, 80 Feet Road, Near Big Bazaar, 6th ...	Spice Elephant	Yes	No	4.1/5	787	Banashankari	Casual Dining	Chinese, North Indian, Thai	800	[('Rated 4.0', 'RATED\n Had been here for din...	[]	Buffet	Banashankari
2	1112, Next to KIMS Medical College, 17th Cross...	San Churro Cafe	Yes	No	3.8/5	918	Banashankari	Cafe, Casual Dining	Cafe, Mexican, Italian	800	[('Rated 3.0', "RATED\n Ambience is not that ...	[]	Buffet	Banashankari
3	1st Floor, Annakuteera, 3rd Stage, Banashankar...	Addhuri Udupi Bhojana	No	No	3.7/5	88	Banashankari	Quick Bites	South Indian, North Indian	300	[('Rated 4.0', "RATED\n Great food and proper...	[]	Buffet	Banashankari
4	10, 3rd Floor, Lakshmi Associates, Gandhi Baza...	Grand Village	No	No	3.8/5	166	Basavanagudi	Casual Dining	North Indian, Rajasthani	600	[('Rated 4.0', 'RATED\n Very good restaurant ...	[]	Buffet	Banashankari

df.shape

(51717, 14)

# checking the available features after dropping the unnecessary features
df.columns

Index(['address', 'name', 'online_order', 'book_table', 'rate', 'votes',
       'location', 'rest_type', 'cuisines', 'approx_cost', 'reviews_list',
       'menu_item', 'type', 'listed_city'],
      dtype='object')

'rate' column has 15% missing values and also its is in string format which is not ideal. Hence we will convert rate into float format

# checking unique values in rate column
df.rate.unique()

array(['4.1/5', '3.8/5', '3.7/5', '3.6/5', '4.6/5', '4.0/5', '4.2/5',
       '3.9/5', '3.1/5', '3.0/5', '3.2/5', '3.3/5', '2.8/5', '4.4/5',
       '4.3/5', 'NEW', '2.9/5', '3.5/5', nan, '2.6/5', '3.8 /5', '3.4/5',
       '4.5/5', '2.5/5', '2.7/5', '4.7/5', '2.4/5', '2.2/5', '2.3/5',
       '3.4 /5', '-', '3.6 /5', '4.8/5', '3.9 /5', '4.2 /5', '4.0 /5',
       '4.1 /5', '3.7 /5', '3.1 /5', '2.9 /5', '3.3 /5', '2.8 /5',
       '3.5 /5', '2.7 /5', '2.5 /5', '3.2 /5', '2.6 /5', '4.5 /5',
       '4.3 /5', '4.4 /5', '4.9/5', '2.1/5', '2.0/5', '1.8/5', '4.6 /5',
       '4.9 /5', '3.0 /5', '4.8 /5', '2.3 /5', '4.7 /5', '2.4 /5',
       '2.1 /5', '2.2 /5', '2.0 /5', '1.8 /5'], dtype=object)

We see that there are values like 'NEW', '-', and nan values they cannot be converted to float, we can't remove the data as they might be new restaurants without ratings so we will write a function accept them as exceptions and convert them to nan values

# '/5' needs to be removed as it does not add any value and make readability difficult and we shall create a new column 
# called rating for extracted values

df['rating'] = df['rate'].astype(str).apply(lambda x: x.split('/')[0]) 

# splitting the string at '/' and selecting only the number at 0 index

df.rating

0         4.1
1         4.1
2         3.8
3         3.7
4         3.8
         ... 
51712    3.6 
51713     nan
51714     nan
51715    4.3 
51716    3.4 
Name: rating, Length: 51717, dtype: object

# loop to convert string values to float

while True:
    try:
        df['rating'] = df['rating'].astype(float)
        break
    except ValueError as e:
        error = str(e).split(":")[-1].strip().replace("'", "")
        print(f'Exception occured while converting to float: {error}')
        df['rating'] = df['rating'].apply(lambda x: x.replace(error, str(np.nan)))

Exception occured while converting to float: NEW
Exception occured while converting to float: -

We have converted 'NEW' and '-' values into NaN values

df.head()

	address	name	online_order	book_table	rate	votes	location	rest_type	cuisines	approx_cost	reviews_list	menu_item	type	listed_city	rating
0	942, 21st Main Road, 2nd Stage, Banashankari, ...	Jalsa	Yes	Yes	4.1/5	775	Banashankari	Casual Dining	North Indian, Mughlai, Chinese	800	[('Rated 4.0', 'RATED\n A beautiful place to ...	[]	Buffet	Banashankari	4.1
1	2nd Floor, 80 Feet Road, Near Big Bazaar, 6th ...	Spice Elephant	Yes	No	4.1/5	787	Banashankari	Casual Dining	Chinese, North Indian, Thai	800	[('Rated 4.0', 'RATED\n Had been here for din...	[]	Buffet	Banashankari	4.1
2	1112, Next to KIMS Medical College, 17th Cross...	San Churro Cafe	Yes	No	3.8/5	918	Banashankari	Cafe, Casual Dining	Cafe, Mexican, Italian	800	[('Rated 3.0', "RATED\n Ambience is not that ...	[]	Buffet	Banashankari	3.8
3	1st Floor, Annakuteera, 3rd Stage, Banashankar...	Addhuri Udupi Bhojana	No	No	3.7/5	88	Banashankari	Quick Bites	South Indian, North Indian	300	[('Rated 4.0', "RATED\n Great food and proper...	[]	Buffet	Banashankari	3.7
4	10, 3rd Floor, Lakshmi Associates, Gandhi Baza...	Grand Village	No	No	3.8/5	166	Basavanagudi	Casual Dining	North Indian, Rajasthani	600	[('Rated 4.0', 'RATED\n Very good restaurant ...	[]	Buffet	Banashankari	3.8

df.rating.dtypes # rating has been  successfully converted to float

dtype('float64')

#droping old rate column which is no more necessary
df.drop('rate', axis = 1, inplace = True)

df.head()

	address	name	online_order	book_table	votes	location	rest_type	cuisines	approx_cost	reviews_list	menu_item	type	listed_city	rating
0	942, 21st Main Road, 2nd Stage, Banashankari, ...	Jalsa	Yes	Yes	775	Banashankari	Casual Dining	North Indian, Mughlai, Chinese	800	[('Rated 4.0', 'RATED\n A beautiful place to ...	[]	Buffet	Banashankari	4.1
1	2nd Floor, 80 Feet Road, Near Big Bazaar, 6th ...	Spice Elephant	Yes	No	787	Banashankari	Casual Dining	Chinese, North Indian, Thai	800	[('Rated 4.0', 'RATED\n Had been here for din...	[]	Buffet	Banashankari	4.1
2	1112, Next to KIMS Medical College, 17th Cross...	San Churro Cafe	Yes	No	918	Banashankari	Cafe, Casual Dining	Cafe, Mexican, Italian	800	[('Rated 3.0', "RATED\n Ambience is not that ...	[]	Buffet	Banashankari	3.8
3	1st Floor, Annakuteera, 3rd Stage, Banashankar...	Addhuri Udupi Bhojana	No	No	88	Banashankari	Quick Bites	South Indian, North Indian	300	[('Rated 4.0', "RATED\n Great food and proper...	[]	Buffet	Banashankari	3.7
4	10, 3rd Floor, Lakshmi Associates, Gandhi Baza...	Grand Village	No	No	166	Basavanagudi	Casual Dining	North Indian, Rajasthani	600	[('Rated 4.0', 'RATED\n Very good restaurant ...	[]	Buffet	Banashankari	3.8

# datatype of approx_cost is also object which is not the right datatype 
df.approx_cost.dtypes

dtype('O')

#Changing the datatype of approx_cost column from object to float64

df['approx_cost'] = df['approx_cost'].astype(str) #to convert to string object to replace ','
df['approx_cost'] = df['approx_cost'].apply(lambda x: x.replace(',', '')) # removing ',' from the string
df['approx_cost'] = df['approx_cost'].astype(float) # converting to float64

df.approx_cost.dtypes

dtype('float64')

EDA

Visualizing restaurant distribution by location

viz2 = go.Bar(
                x = df.location.value_counts().keys(),
                y = df.location.value_counts(),
                name = "location",
                text = None)
data2 = [viz2]
layout = go.Layout(title = 'Restaurant Distribution by Location', 
                   barmode = "group", 
                   yaxis=dict(title= 'Number of Restaurants'))
fig = go.Figure(data = data2, layout = layout)
iplot(fig, filename = 'plot2')

We see that 'BTM' has the highest number of restaurants followed by 'HSR' and 'Koramangala' all are know for their closeness to IT hubs and Outer ringroad. we can say IT industry has been the driving force of restaurant industry in Benagaluru

Restaurant distibution by type of service offered

viz1 = go.Bar(
                x = df['type'].value_counts().keys(),
                y = df['type'].value_counts(),
                name = "type",
                text = None)
data1 = [viz1]
layout = go.Layout(title = 'Restaurant Distribution by Type of Service Offered', 
                   barmode = "group", 
                   yaxis=dict(title= 'Number of Restaurants'))
fig = go.Figure(data = data1, layout = layout)
iplot(fig, filename = 'plot_1')

We can see that majority restaurants are supporting food delivery, which is an indication of growing online trend

Restaurant distribution by subcategories

viz3 = go.Bar(
                x = df['rest_type'].value_counts().head(15).keys(),
                y = df['rest_type'].value_counts().head(15),
                name = "rest_type",
                text = None)
data3 = [viz3]
layout = go.Layout(title = 'Restaurant Distribution by Sub-Categories', 
                   barmode = "group", 
                   yaxis=dict(title= 'Number of Restaurants'))
fig = go.Figure(data = data3, layout = layout)
iplot(fig, filename = 'plot3')

'Quick-bites' category has the majority share, which indicates growing fast food culture in the city

How many of these restaurants accept online orders ?

plt.figure(figsize=(10,10))
x = df['online_order'].value_counts()
plt.pie(x,labels=['Accepted','Not Accepted'],autopct='%1.1f%%',textprops = {'fontsize': 16}, colors = ['#99ff99', '#ff8080'])
circle = plt.Circle( (0,0), 0.8, color='white')
p=plt.gcf()
p.gca().add_artist(circle)

plt.title("Online Order Service Offered", fontsize = 25)

plt.show()

We can see almost 58% of restaurants are accepting online orders. There is still around 40% restaurants that have not yet adopted online delivery service. So there is still potential for online delivery service service companies to penetrate the market

How many of these restaurants offer dine - in service ?

plt.figure(figsize=(10,10))
x = df['book_table'].value_counts()
plt.pie(x,labels=['Yes','No'],autopct='%1.1f%%',textprops = {'fontsize': 16}, colors = ['darkslateblue', 'lightsalmon'])
circle = plt.Circle( (0,0), 0.8, color='white')
p=plt.gcf()
p.gca().add_artist(circle)

plt.title("Dine-In Service Offered", fontsize = 25)

plt.show()

We can see 87.5% of the restaurants offer Dine in, it can be conluded by comparing above two charts that majority restaurants are offering both online and dine in service. and only about 12.5% of restaurant are online service. They may represent cloud kitchens.

We can see that there is a huge market opportunity in cloud kitchen service still untapped. and competion is not so much in this segment yet

Top 10 restaurant chains based on number of outlets

restaurant_chains = df['name'].value_counts().to_frame()
restaurant_chains['Restaurant Name'] = restaurant_chains.index
restaurant_chains = restaurant_chains.reset_index(drop = True)
restaurant_chains.rename(columns={'name':'Count'}, inplace=True)
restaurant_chains.reset_index(drop = True)
restaurant_chains = restaurant_chains.loc[:, ['Restaurant Name', 'Count']]
restaurant_chains.head(10)

	Restaurant Name	Count
0	Cafe Coffee Day	96
1	Onesta	85
2	Just Bake	73
3	Empire Restaurant	71
4	Five Star Chicken	70
5	Kanti Sweets	68
6	Petoo	66
7	Polar Bear	65
8	Baskin Robbins	64
9	Chef Baker's	62

'Cafe coffee day' has the highest number of outlets followed by 'Onesta' and 'Just Bake', this shows the love for Bengalureans towards coffee, pizza and deserts

px.bar(restaurant_chains.head(10) , x='Restaurant Name', y='Count', title = "Top Restaurant Chains In Bengaluru")

Top restaurants in quick-bite category

quick_bite = df[df['rest_type'] == 'Quick Bites'].name.value_counts().head(10).to_frame()
quick_bite['Restaurant Name'] = quick_bite.index
quick_bite = quick_bite.reset_index(drop = True)
quick_bite.rename(columns={'name':'Count'}, inplace=True)
quick_bite.reset_index(drop = True)
quick_bite = quick_bite.loc[:, ['Restaurant Name', 'Count']]
quick_bite

	Restaurant Name	Count
0	Five Star Chicken	69
1	Domino's Pizza	60
2	McDonald's	59
3	KFC	56
4	Ambur Hot Dum Biryani	53
5	Rolls On Wheels	51
6	Burger King	51
7	Pizza Stop	48
8	Goli Vada Pav No. 1	46
9	Subway	45

px.bar(quick_bite, x='Restaurant Name', y='Count', title = "Top Quick Bite Places")

Looks like 'five star chicken' is leading the quick-bites category followed by "Domino's" and "McDonald's"

looks like multi national food service gaints are leading this category. Domino's which is an indian company stands second in the segment

Top restuarants in casual dining category

casual_dining = df[df['rest_type'] == 'Casual Dining'].name.value_counts().head(10).to_frame()
casual_dining['Restaurant Name'] = casual_dining.index
casual_dining = casual_dining.reset_index(drop = True)
casual_dining.rename(columns={'name':'Count'}, inplace=True)
casual_dining.reset_index(drop = True)
casual_dining = casual_dining.loc[:, ['Restaurant Name', 'Count']]
casual_dining

	Restaurant Name	Count
0	Empire Restaurant	58
1	Beijing Bites	48
2	Mani's Dum Biryani	47
3	Chung Wah	46
4	Barbeque Nation	41
5	Toscano	41
6	Oye Amritsar	41
7	A2B - Adyar Ananda Bhavan	39
8	New Prashanth Hotel	38
9	Pizza Hut	38

px.bar(casual_dining, x='Restaurant Name', y='Count', title = "Top Casual Dining Places")

'Empire Restaurant', 'Beijing Bites', 'Mani's Dum Biryani' are the top 3 restaurants in casual dining category. this shows the Bengalureans love towards biriyani and chinese cusinies

Top 10 restuarants based on rating

popular_franchises = df.groupby(by='name', as_index=False).agg({'votes': 'sum','address': 'count','approx_cost': 'mean', 'rating': 'mean'})                                                               
popular_franchises.columns = ['name', 'total_votes', 'total_units', 'mean_approx_cost', 'mean_rating'] #renaming the columns appropriately
popular_franchises = popular_franchises.sort_values(by='mean_rating', ascending=False)

# organising the dataframe
popular_franchises = popular_franchises.loc[:, ['name', 'total_units', 'total_votes',
                                                'mean_approx_cost', 'mean_rating']]

popular_franchises.head()

	name	total_units	total_votes	mean_approx_cost	mean_rating
597	Asia Kitchen By Mainland China	19	42273	1500.0	4.900000
1274	Byg Brewski Brewing Company	6	99531	1600.0	4.900000
6552	SantÃÃÃ...	1	246	1000.0	4.900000
5927	Punjab Grill	7	9660	2000.0	4.871429
865	Belgian Waffle Factory	29	24882	400.0	4.844828

There seems to be a error in name of the restaurant, lets try to correct it

# locating the name with the index
popular_franchises.loc[6552]['name']

'SantÃ\x83Â\x83Ã\x82Â\x83Ã\x83Â\x82Ã\x82Â\x83Ã\x83Â\x83Ã\x82Â\x82Ã\x83Â\x82Ã\x82Â\x83Ã\x83Â\x83Ã\x82Â\x83Ã\x83Â\x82Ã\x82Â\x82Ã\x83Â\x83Ã\x82Â\x82Ã\x83Â\x82Ã\x82© Spa Cuisine'

Looks like original name is 'Santa Spa Cuisine'

# lets replace with the correct name
a = 'SantÃ\x83Â\x83Ã\x82Â\x83Ã\x83Â\x82Ã\x82Â\x83Ã\x83Â\x83Ã\x82Â\x82Ã\x83Â\x82Ã\x82Â\x83Ã\x83Â\x83Ã\x82Â\x83Ã\x83Â\x82Ã\x82Â\x82Ã\x83Â\x83Ã\x82Â\x82Ã\x83Â\x82Ã\x82© Spa Cuisine'
popular_franchises['name'] = popular_franchises['name'].apply(lambda x: 'Santa Spa Cusisine' if x == a else x)

# top 10 highest rated restaurants 
print("Top 10 Highest Rated Restaurant Chains")
popular_franchises.head(10)

Top 10 Highest Rated Restaurant Chains

	name	total_units	total_votes	mean_approx_cost	mean_rating
597	Asia Kitchen By Mainland China	19	42273	1500.000000	4.900000
1274	Byg Brewski Brewing Company	6	99531	1600.000000	4.900000
6552	Santa Spa Cusisine	1	246	1000.000000	4.900000
5927	Punjab Grill	7	9660	2000.000000	4.871429
865	Belgian Waffle Factory	29	24882	400.000000	4.844828
2598	Flechazo	6	29956	1400.000000	4.800000
5471	O.G. Variar & Sons	2	2317	200.000000	4.800000
8035	The Pizza Bakery	6	10523	1200.000000	4.800000
129	AB's - Absolute Barbecues	23	86418	1568.421053	4.789474
964	Biergarten	18	39843	2183.333333	4.766667

10 Lowest rated Restaurant Chains

not_popular_franchises = popular_franchises.sort_values(by='mean_rating', ascending = True)

print("10 Lowest rated Restaurant Chains")
not_popular_franchises.head(10)

10 Lowest rated Restaurant Chains

	name	total_units	total_votes	mean_approx_cost	mean_rating
316	Alibi - Maya International Hotel	5	1123	1200.0	1.800000
2778	Fusion Lounge	9	3573	1500.0	2.000000
2112	Decker's Lane	4	970	400.0	2.100000
705	Bageecha	4	1916	650.0	2.150000
4636	Mamma Mexicana	13	5296	1000.0	2.200000
7584	Taste Of Kerala	5	240	600.0	2.240000
911	Bhagini	12	1683	800.0	2.283333
4788	Meghana Biryani	5	1185	800.0	2.300000
1044	Biryani Junction	9	2682	450.0	2.300000
8422	Vande Matharam	4	716	500.0	2.300000

Most popular cuisines

cuisines = df['cuisines'].value_counts().head(20).to_frame()
cuisines['Cuisine names'] = cuisines.index
cuisines = cuisines.reset_index(drop = True)
cuisines.rename(columns={'cuisines':'Count'}, inplace=True)
cuisines = cuisines.loc[:, ['Cuisine names', 'Count']]
cuisines

	Cuisine names	Count
0	North Indian	2913
1	North Indian, Chinese	2385
2	South Indian	1828
3	Biryani	918
4	Bakery, Desserts	911
5	Fast Food	803
6	Desserts	766
7	Cafe	756
8	South Indian, North Indian, Chinese	726
9	Bakery	651
10	Chinese	556
11	Ice Cream, Desserts	417
12	Chinese, North Indian	415
13	Mithai, Street Food	372
14	Desserts, Ice Cream	354
15	North Indian, Chinese, Biryani	352
16	North Indian, South Indian	343
17	South Indian, North Indian	343
18	North Indian, South Indian, Chinese	305
19	Beverages	301

px.bar(cuisines, x="Cuisine names", y="Count", title = "Top Cuisines Bengalureans Prefer")

Distribution of rating

plt.figure(figsize=(10,7))
rating = df['rating'].value_counts()
sns.barplot(x= rating.index,y=rating)
plt.xlabel("Ratings", fontsize = 12)
plt.ylabel('Count',fontsize = 12)
plt.title("Distribution of Ratings", fontsize = 18)

plt.show()

From the visualization we can infer 3.9 is most common rating

Distribution of cost for 2 people

plt.figure(figsize=(10,8))
sns.histplot(df.approx_cost,bins = 50,kde = False)
plt.title("Distribution Plot of Approximate Cost", fontsize = 15)
plt.xlabel("Approximate Cost for 2 People",fontsize = 12)
plt.ylabel('Count',fontsize = 12)
plt.show()

From the above plot we can conclude it is going to cost an approximate 500/- INR for two people in majority of the restaurants in Bengaluru

Best budget friendly restaurants

df_cheap = df[['name','approx_cost','location','rest_type','cuisines']].groupby(['approx_cost'], sort = True)
df_cheap = df_cheap.filter(lambda x: x.mean() <= 1500)
df_cheap = df_cheap.sort_values(by=['approx_cost'])
df_cheap

	name	approx_cost	location	rest_type	cuisines
18891	Srinidhi Sagar Food Line	40.0	Indiranagar	Quick Bites	South Indian, North Indian, Chinese
32485	Srinidhi Sagar	40.0	Old Airport Road	Quick Bites	South Indian, North Indian, Chinese
5270	Srinidhi Sagar Food Line	40.0	Indiranagar	Quick Bites	South Indian, North Indian, Chinese
14819	Srinidhi Sagar Food Line	40.0	Indiranagar	Quick Bites	South Indian, North Indian, Chinese
27091	Srinidhi Sagar Deluxe	40.0	Domlur	Quick Bites	South Indian, North Indian, Chinese
...	...	...	...	...	...
5516	Feast India Co.	1500.0	Cunningham Road	Casual Dining	Awadhi, Bengali, North Indian
19279	Barebones - The Balcony Bar	1500.0	Indiranagar	Bar, Casual Dining	Chinese, Continental, Finger Food, Italian
19278	Lono	1500.0	Indiranagar	Lounge, Casual Dining	Continental, Modern Indian
6103	Chianti	1500.0	MG Road	Casual Dining	Italian
51716	The Nest - The Den Bengaluru	1500.0	ITPL Main Road, Whitefield	Bar, Casual Dining	Finger Food, North Indian, Continental

49592 rows × 5 columns

Best restaurants for fine dining

df_expen = df[['name','approx_cost','location','rest_type','cuisines']].groupby(['approx_cost'])
df_expen = df_expen.filter(lambda x: x.mean() >= 3000)
df_expen = df_expen.sort_values(by=['approx_cost'])
df_expen

	name	approx_cost	location	rest_type	cuisines
4983	Rim Naam - The Oberoi	3000.0	MG Road	Fine Dining	Thai
39120	Ssaffron - Shangri-La Hotel	3000.0	Vasanth Nagar	Fine Dining, Bar	North Indian
39127	Hype - Shangri-La Hotel	3000.0	Vasanth Nagar	Lounge, Bar	Finger Food
39138	Mikusu - Conrad Bengaluru	3000.0	Ulsoor	Fine Dining	Japanese, Chinese, Thai
39150	Indian Durbar - Conrad Bengaluru	3000.0	Ulsoor	Fine Dining	North Indian
...	...	...	...	...	...
42141	Malties - Radisson Blu	4500.0	Marathahalli	Lounge	Continental, Fast Food
41591	Malties - Radisson Blu	4500.0	Marathahalli	Lounge	Continental, Fast Food
40266	Royal Afghan - ITC Windsor	5000.0	Sankey Road	Fine Dining	North Indian, Mughlai
45618	Le Cirque Signature - The Leela Palace	6000.0	Old Airport Road	Fine Dining	French, Italian
19139	Le Cirque Signature - The Leela Palace	6000.0	Old Airport Road	Fine Dining	French, Italian

241 rows × 5 columns

Top rated restaurants

df_rating = df[['name','rating']].groupby(['rating']) 
df_rating = df_rating.filter(lambda x: x.mean() >= 4.5) # above returns a groupby object hence using labmda function
df_rating = df_rating.sort_values(by=['rating'],ascending = False)
df_rating.head(10)

	name	rating
43055	Belgian Waffle Factory	4.9
49170	Byg Brewski Brewing Company	4.9
50059	Byg Brewski Brewing Company	4.9
3921	Byg Brewski Brewing Company	4.9
42381	Belgian Waffle Factory	4.9
36684	Asia Kitchen By Mainland China	4.9
11504	Asia Kitchen By Mainland China	4.9
4801	Byg Brewski Brewing Company	4.9
9099	Asia Kitchen By Mainland China	4.9
49627	Byg Brewski Brewing Company	4.9

Finest taste on budget

cheap_and_best = pd.merge(df_cheap, df_rating, how='inner', on=['name'])

expensive_and_best = pd.merge(df_expen, df_rating, how='inner', on=['name'])

print("Top 10 High Rated Restaurants When under Budget")
cheap_and_best.head(10)

Top 10 High Rated Restaurants When under Budget

	name	approx_cost	location	rest_type	cuisines	rating
0	Brahmin's Coffee Bar	100.0	Basavanagudi	Quick Bites	South Indian	4.8
1	Brahmin's Coffee Bar	250.0	Malleshwaram	Quick Bites	South Indian	4.8
2	Taaza Thindi	100.0	Banashankari	Quick Bites	South Indian	4.7
3	Natural Ice Cream	150.0	Jayanagar	Dessert Parlor	Ice Cream, Desserts	4.6
4	Natural Ice Cream	150.0	Jayanagar	Dessert Parlor	Ice Cream, Desserts	4.6
5	Natural Ice Cream	150.0	Jayanagar	Dessert Parlor	Ice Cream, Desserts	4.5
6	Natural Ice Cream	150.0	Jayanagar	Dessert Parlor	Ice Cream, Desserts	4.5
7	Natural Ice Cream	150.0	Jayanagar	Dessert Parlor	Ice Cream, Desserts	4.5
8	Natural Ice Cream	150.0	Jayanagar	Dessert Parlor	Ice Cream, Desserts	4.5
9	Natural Ice Cream	150.0	Jayanagar	Dessert Parlor	Ice Cream, Desserts	4.5

Top high rated restaurants when budget is not a issue

print("Top 10 High Rated Restaurants For Fine Dining")
expensive_and_best.value_counts()

Top 10 High Rated Restaurants For Fine Dining

name                           approx_cost  location        rest_type    cuisines                              rating
Rim Naam - The Oberoi          3000.0       MG Road         Fine Dining  Thai                                  4.6       144
Karavalli - The Gateway Hotel  3500.0       Residency Road  Fine Dining  Mangalorean, Konkan, Seafood, Kerala  4.5        25
Alba - JW Marriott Bengaluru   4000.0       Lavelle Road    Fine Dining  Italian                               4.5         4
dtype: int64

Locations where you can find best restaurants when on budget

  top_cheap_locations = cheap_and_best["location"].value_counts().to_frame()
top_cheap_locations = top_cheap_locations.reset_index(level=0)
top_cheap_locations.rename(columns={'index':'location', 'location': 'Count'}, inplace=True)
top_cheap_locations.head(5)

	location	Count
0	Koramangala 5th Block	5142
1	Cunningham Road	1934
2	Indiranagar	1290
3	Koramangala 7th Block	1161
4	BTM	916

px.bar(top_cheap_locations.head(5), x = "location" , y = "Count", title = "Top Locations for Low Price, High Rated Restaurants")

Locations where you can find best restaurants on the expensive side

top_expens_locations = expensive_and_best["location"].value_counts().to_frame()
top_expens_locations = top_expens_locations.reset_index(level=0)
top_expens_locations.rename(columns={'index':'location', 'location': 'Count'}, inplace=True)
top_expens_locations.head(10)

	location	Count
0	MG Road	144
1	Residency Road	25
2	Lavelle Road	4

px.bar(top_expens_locations.head(10), x = 'location' , y = "Count", title = "Top Locations for Expensive, High Rated Restaurants")

location_cost = df.groupby(by='location', as_index=False).agg({'approx_cost': 'mean'})                                                               
location_cost.columns = ['location', 'mean_approx_cost']
location_cost = location_cost.sort_values(by='mean_approx_cost', ascending=False)

location_cost.head(10)

	location	mean_approx_cost
75	Sankey Road	2505.555556
66	Race Course Road	1309.352518
51	Lavelle Road	1307.934990
52	MG Road	1155.704698
28	Infantry Road	1062.251656
70	Residency Road	966.320475
50	Langford Town	883.333333
81	St. Marks Road	871.306818
15	Cunningham Road	864.969450
12	Church Street	834.885764

px.bar(location_cost, x = 'location' , y = "mean_approx_cost", title = "Approximate Cost By Location",  labels = {'location' : 'Location', "mean_approx_cost" : 'Approximate cost for 2 people'} )

Lets Visualize the distribution of restaurants in Bengaluru

from geopy.geocoders import Nominatim
import folium
from folium.plugins import HeatMap
from folium.plugins import FastMarkerCluster

len(df['location'].unique())

94

locations = pd.DataFrame({"Name" : df.location.unique()})

geolocation = Nominatim(user_agent= "app")

locations

	Name
0	Banashankari
1	Basavanagudi
2	Mysore Road
3	Jayanagar
4	Kumaraswamy Layout
...	...
89	West Bangalore
90	Magadi Road
91	Yelahanka
92	Sahakara Nagar
93	Peenya

94 rows × 1 columns

lat=[]
lon=[]
for i in locations['Name']:
    i = geolocation.geocode(i)    
    if i is None:
        lat.append(np.nan)
        lon.append(np.nan)
    else:
        lat.append(i.latitude)
        lon.append(i.longitude)

locations["latitude"] = lat
locations["longitude"] = lon

locations.head()

	Name	latitude	longitude
0	Banashankari	15.887678	75.704678
1	Basavanagudi	12.941726	77.575502
2	Mysore Road	12.387214	76.666963
3	Jayanagar	27.643927	83.052805
4	Kumaraswamy Layout	12.908149	77.555318

Restaurant_locations = pd.DataFrame(df['location'].value_counts().reset_index())
Restaurant_locations.columns = ['Name','count']
Restaurant_locations.head()

	Name	count
0	BTM	5124
1	HSR	2523
2	Koramangala 5th Block	2504
3	JP Nagar	2235
4	Whitefield	2144

print(locations.shape)
print(Restaurant_locations.shape)

(94, 3)
(93, 2)

def generateBaseMap(default_location=[12.97, 77.59], default_zoom_start=12):
    base_map = folium.Map(location=default_location, zoom_start=default_zoom_start)
    return base_map

basemap = generateBaseMap()
print("Base Map of Bengaluru")
basemap

Base Map of Bengaluru

Make this Notebook Trusted to load map: File -> Trust Notebook

Restaurant_locations = Restaurant_locations.merge(locations,on='Name',how="left").dropna()
Restaurant_locations.head()

	Name	count	latitude	longitude
0	BTM	5124	45.954851	-112.496595
1	HSR	2523	18.147500	41.538889
2	Koramangala 5th Block	2504	12.934377	77.628415
3	JP Nagar	2235	12.265594	76.646540
4	Whitefield	2144	44.373058	-71.611858

HeatMap(Restaurant_locations[['latitude','longitude','count']],zoom = 25,radius = 20).add_to(basemap)
print("Heatmap of Restaurants Distribution in Bengaluru")
basemap

Heatmap of Restaurants Distribution in Bengaluru

Make this Notebook Trusted to load map: File -> Trust Notebook

We can see that most restaurants are concentrated on the south eastern part of the Bengaluru. Also, there are quite a lot of areas around the central Bengaluru where restaurant concentration are not that high, these might be potential locations for new restaurants looking to capture the market without facing a lot of competition

Sentiment Analysis

from textblob import TextBlob

sentiment = []
for i in range(0,len(df)):
    x = TextBlob(df.loc[i,"reviews_list"])
    if x.sentiment.polarity > 0:
        sentiment.append("positive")
    elif x.sentiment.polarity == 0:
        sentiment.append("neutral")
    else:
        sentiment.append("negative")

df_sent = df.copy()

df_sent["reviews_list"] = sentiment

df_sent.head()

	address	name	online_order	book_table	votes	location	rest_type	cuisines	approx_cost	reviews_list	menu_item	type	listed_city	rating
0	942, 21st Main Road, 2nd Stage, Banashankari, ...	Jalsa	Yes	Yes	775	Banashankari	Casual Dining	North Indian, Mughlai, Chinese	800.0	positive	[]	Buffet	Banashankari	4.1
1	2nd Floor, 80 Feet Road, Near Big Bazaar, 6th ...	Spice Elephant	Yes	No	787	Banashankari	Casual Dining	Chinese, North Indian, Thai	800.0	positive	[]	Buffet	Banashankari	4.1
2	1112, Next to KIMS Medical College, 17th Cross...	San Churro Cafe	Yes	No	918	Banashankari	Cafe, Casual Dining	Cafe, Mexican, Italian	800.0	positive	[]	Buffet	Banashankari	3.8
3	1st Floor, Annakuteera, 3rd Stage, Banashankar...	Addhuri Udupi Bhojana	No	No	88	Banashankari	Quick Bites	South Indian, North Indian	300.0	positive	[]	Buffet	Banashankari	3.7
4	10, 3rd Floor, Lakshmi Associates, Gandhi Baza...	Grand Village	No	No	166	Basavanagudi	Casual Dining	North Indian, Rajasthani	600.0	positive	[]	Buffet	Banashankari	3.8

positive = df_sent[df_sent["reviews_list"] == "positive"]

negative = df_sent[df_sent["reviews_list"] == "negative"]

neutral = df_sent[df_sent["reviews_list"] == "neutral"]

print("Restaurants having positive reviews: ",len(positive))

Restaurants having positive reviews:  39629

print("Restaurants having negative reviews: ",len(negative))

Restaurants having negative reviews:  4324

print("Restaurants having neutral reviews: ",len(neutral))

Restaurants having neutral reviews:  7764

df_sentiment = df_sent["reviews_list"].value_counts().to_frame()

df_sentiment = df_sentiment.reset_index()
df_sentiment = df_sentiment.rename(columns = {'index':'sentiment', 'reviews_list': 'Count'})
df_sentiment

	sentiment	Count
0	positive	39629
1	neutral	7764
2	negative	4324

px.bar(df_sentiment, x = 'sentiment' , y = "Count", title = "Rating Sentiment Distribution", color = 'sentiment',color_discrete_map={'positive':'green',
                                 'neutral':'lightblue',
                                 'negative':'red',})

The data reveals most restaurants are having a positive rating which is a good news for the restaurant industry

Data transformation for model building

df_model = df_sent.copy()

df_model['rating'] = df_model['rating'].replace(np.nan, 0)

df_model.dropna(axis=0, inplace = True)

from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()

# label encoding to to convert categorical variables into the machine-readable form for certain algorithms

x = ['online_order','book_table','rest_type','cuisines','listed_city', 'location', 'type', 'reviews_list']
for i in x:
    df_model[i] = le.fit_transform(df_model[i])

df_model.head()

	address	name	online_order	book_table	votes	location	rest_type	cuisines	approx_cost	reviews_list	menu_item	type	listed_city	rating
0	942, 21st Main Road, 2nd Stage, Banashankari, ...	Jalsa	1	1	775	1	27	2145	800.0	2	[]	0	1	4.1
1	2nd Floor, 80 Feet Road, Near Big Bazaar, 6th ...	Spice Elephant	1	0	787	1	27	947	800.0	2	[]	0	1	4.1
2	1112, Next to KIMS Medical College, 17th Cross...	San Churro Cafe	1	0	918	1	22	761	800.0	2	[]	0	1	3.8
3	1st Floor, Annakuteera, 3rd Stage, Banashankar...	Addhuri Udupi Bhojana	0	0	88	1	78	2539	300.0	2	[]	0	1	3.7
4	10, 3rd Floor, Lakshmi Associates, Gandhi Baza...	Grand Village	0	0	166	4	27	2174	600.0	2	[]	0	1	3.8

df_model.drop(columns = ['address','menu_item', 'name'], axis = 1, inplace = True)

df_model.shape

(51148, 11)

features1 = df_model.drop(['rating'], axis = 1)

target = df_model['rating']

# Scaling the features to bring various features under a similar scale and improve model performance
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()

features = scaler.fit_transform(features1)

features

array([[ 0.82912014,  2.63270813,  0.60767491, ...,  0.51104057,
        -2.46134804, -1.59071902],
       [ 0.82912014, -0.37983702,  0.62255425, ...,  0.51104057,
        -2.46134804, -1.59071902],
       [ 0.82912014, -0.37983702,  0.78498706, ...,  0.51104057,
        -2.46134804, -1.59071902],
       ...,
       [-1.20609783, -0.37983702, -0.35328253, ..., -1.1055531 ,
         2.79876977,  1.78898279],
       [-1.20609783,  2.63270813, -0.06065549, ...,  0.51104057,
         2.79876977,  1.78898279],
       [-1.20609783, -0.37983702, -0.33716325, ...,  0.51104057,
         2.79876977,  1.78898279]])

Let's split the data into training and test set. We will be using 80% of data for training and 20% of data as test set

from sklearn.model_selection import train_test_split
x_train, x_test, y_train, y_test = train_test_split(features , target, test_size=0.20, random_state=123)

# Check the shape of splited data
x_train.shape,x_test.shape,y_train.shape,y_test.shape

((40918, 10), (10230, 10), (40918,), (10230,))

Model building

Linear Regression

from sklearn.linear_model import LinearRegression

lr = LinearRegression()
lr.fit(x_train,y_train)
y_predict_lr = lr.predict(x_test) 

from sklearn.metrics import r2_score
r2_score(y_test,y_predict_lr)

0.3415262784393306

The model accuracy for linear regression model is very low. Hyperparameter Tuning techniques like Lasso and Ridge may not improve the model performance drastically. Hence, lets check out other model and decide on the hyper parameter tuning

Decision Tree

from sklearn.tree import DecisionTreeRegressor

dtree = DecisionTreeRegressor()
dtree.fit(x_train,y_train)
y_predict_dt = dtree.predict(x_test)

r2_score(y_test,y_predict_dt)

0.9808213833054206

The tree based model is performing well in terms of R squared score. lets us use an ensemble learning to boost the performance further

Random Forest

#Preparing Random Forest Regression
from sklearn.ensemble import RandomForestRegressor
r_forest = RandomForestRegressor(n_estimators=200, random_state = 123)
r_forest.fit(x_train,y_train)
y_predict_rf = r_forest.predict(x_test)

r2_score(y_test,y_predict_rf)

0.9896641961151734

Random Forest Regressor is performing better than decision tree regressor. lets check out other models as well

Gradient Boosting

from sklearn.ensemble import GradientBoostingRegressor

gbr = GradientBoostingRegressor()
gbr.fit(x_train, y_train)

y_predict_gbr = gbr.predict(x_test)

r2_score(y_test,y_predict_gbr)

0.9623416944579003

Performance of gradient boosting algorithm has degraded a little compared to random forest regressor. lets checkout Xtreme gradient boosting.

XG boost

from xgboost import XGBRegressor

xgbr = XGBRegressor(booster = 'gbtree', learning_rate = 0.1, max_depth = 7, n_estimators = 200)
xgbr.fit(x_train, y_train)
y_predict_xgbr = xgbr.predict(x_test)

r2_score(y_test,y_predict_xgbr)

0.9768642352075912

all though XG boost regressor performed well compared to gradient boosting its score is still lesser than random forest regressor

Tabulating the model results

models = pd.DataFrame({
    'Model' : ['Linear Regression', 'Decision Tree', 'Random Forest', 'Gradient Boost', 'XgBoost'],
    'R2_Score' : [lr.score(x_test, y_test), dtree.score(x_test, y_test), r_forest.score(x_test, y_test), gbr.score(x_test, y_test), xgbr.score(x_test, y_test)]
})


models.sort_values(by = 'R2_Score', ascending = False)

	Model	R2_Score
2	Random Forest	0.989664
1	Decision Tree	0.980821
4	XgBoost	0.976864
3	Gradient Boost	0.962342
0	Linear Regression	0.341526

As we can see Random Forest Regressor seem to be the best performing model. we will save this model for making further predictions

Model Prediction

prediction_df = pd.DataFrame({"Actual Rating": y_test, "Predicted Rating": np.round(y_predict_rf,1)})

random_state = 43
prediction_df.sample(10)

	Actual Rating	Predicted Rating
37709	3.3	3.3
10420	4.1	4.0
7434	3.2	3.7
46091	0.0	0.0
40185	4.0	3.8
30663	0.0	0.0
26577	4.0	4.0
41542	4.6	4.6
46491	3.2	3.2
18955	3.3	3.3

Saving the model for deployment

import pickle

filename = 'final_model.sav'
pickle.dump(r_forest, open(filename, 'wb'))

loaded_model = pickle.load(open(filename, 'rb'))
result = loaded_model.score(x_test, y_test)
print(result)

0.9896641961151734

Recommender system

# creating a copy of the data
df_recommender = df.sample(frac=0.5)

df_recommender.shape

(25858, 14)

# droppping unwanted columns
df_recommender = df_recommender.drop(['address','rest_type', 'type', 'menu_item', 'votes'],axis=1)

df_recommender.head()

	name	online_order	book_table	location	cuisines	approx_cost	reviews_list	listed_city	rating
49709	Punjabi Last Stand	No	No	Sarjapur Road	North Indian	300.0	[]	Sarjapur Road	NaN
2780	Thyme & Whisk	Yes	No	Jayanagar	Asian, Chinese, Continental, Italian	800.0	[('Rated 2.0', 'RATED\n For starters, we orde...	Basavanagudi	NaN
13503	Sri Krishna Sagar	Yes	No	Electronic City	South Indian, North Indian	300.0	[('Rated 5.0', 'RATED\n Tasty food and good a...	Electronic City	3.5
26424	Ambur Khousia Biryani Corner	No	No	Banaswadi	Biryani	100.0	[]	Kammanahalli	NaN
6242	Strikers Sports Bar and Music	No	Yes	Shanti Nagar	Continental, North Indian, Chinese	1200.0	[('Rated 1.0', "RATED\n Dikhave Pe Mat Jao, A...	Brigade Road	3.8

df_recommender.set_index('name', inplace=True)

indices = pd.Series(df_recommender.index)

indices

0                   Punjabi Last Stand
1                        Thyme & Whisk
2                    Sri Krishna Sagar
3         Ambur Khousia Biryani Corner
4        Strikers Sports Bar and Music
                     ...              
25853     Calcutta Victoria Chat House
25854                   Cake N Cookies
25855                      Burger King
25856                        Momo Time
25857                        Miss Momo
Name: name, Length: 25858, dtype: object

import re
from nltk.corpus import stopwords
from sklearn.metrics.pairwise import linear_kernel
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.feature_extraction.text import TfidfVectorizer

tfidf = TfidfVectorizer(analyzer='word', ngram_range=(1, 2), min_df=0, stop_words='english')
tfidf_matrix = tfidf.fit_transform(df_recommender['reviews_list'])

cosine_similarities = linear_kernel(tfidf_matrix, tfidf_matrix)

def recommend(name, cosine_similarities = cosine_similarities):
    
    recommend_restaurant = []  # Create a list to put top restaurants
    idx = indices[indices == name].index[0] # Find the index of the hotel entered
    # Find the restaurants with a similar cosine-sim value and order them from biggest number
    score_series = pd.Series(cosine_similarities[idx]).sort_values(ascending=False)
    
    # Extract top 30 restaurant indexes with a similar cosine-sim value
    top30_indexes = list(score_series.iloc[0:31].index)
    
    # Names of the top 30 restaurants
    for each in top30_indexes:
        recommend_restaurant.append(list(df_recommender.index)[each])
    
    # Creating the new data set to show similar restaurants
    df_new = pd.DataFrame(columns=['cuisines', 'rating', 'approx_cost'])
    
    # Create the top 30 similar restaurants with some of their columns
    for each in recommend_restaurant:
        df_new = df_new.append(pd.DataFrame(df_recommender[['cuisines','rating', 'approx_cost', 'location']][df_recommender.index == each].sample()))
    
    # Drop the same named restaurants and sort only the top 10 by the highest rating
    df_new = df_new.drop_duplicates(subset=['cuisines', 'rating', 'approx_cost','location'], keep=False)
    df_new = df_new.sort_values(by='rating', ascending=False).head(10)
     
    print('TOP %s RESTAURANTS LIKE %s WITH SIMILAR REVIEWS: ' % (str(len(df_new)), name))
    return df_new

def search_restaurant(restaurant_name):
    x = df['name'].str.contains(restaurant_name, case = False, na = False)
    return df[x].head()

Search for a restaurant by name

search_restaurant('Absolute')

	address	name	online_order	book_table	votes	location	rest_type	cuisines	approx_cost	reviews_list	menu_item	type	listed_city	rating
3765	1st Floor, GRS Towers, Above Spencers Hyper Ma...	AB's - Absolute Barbecues	No	Yes	881	Sarjapur Road	Casual Dining	European, Mediterranean, North Indian, BBQ	1600.0	[('Rated 2.0', "RATED\n Looks AB is finding h...	[]	Buffet	Bellandur	4.7
4141	Behind Prestige Tech Park, Kadubisanahalli, Ma...	Absolute Shawarma	Yes	No	116	Marathahalli	Takeaway, Delivery	Lebanese	150.0	[('Rated 4.0', 'RATED\n Good food at a budget...	[]	Delivery	Bellandur	3.8
4809	1st Floor, GRS Towers, Above Spencers Hyper Ma...	AB's - Absolute Barbecues	No	Yes	881	Sarjapur Road	Casual Dining	European, Mediterranean, North Indian, BBQ	1600.0	[('Rated 2.0', "RATED\n Looks AB is finding h...	[]	Dine-out	Bellandur	4.7
6745	2nd Floor, I20-A2, EPIP Zone, Near Vydehi Hosp...	AB's - Absolute Barbecues	No	Yes	2882	Whitefield	Casual Dining	European, Mediterranean, North Indian, BBQ	1600.0	[('Rated 4.0', 'RATED\n Went today for Lunch ...	[]	Buffet	Brookefield	4.8
7601	750/1, Near Salem Kitchen, ITPL Main Road, Kun...	Absolute Shawarma	Yes	No	0	Brookefield	Takeaway, Delivery	Lebanese	150.0	[('Rated 5.0', 'RATED\n The shawarma was real...	['Arabian Combo', 'Mexican Combo', 'Arabian Sh...	Delivery	Brookefield	NaN

Get recommendation of similar restaurants restaurants

recommend("AB's - Absolute Barbecues")

TOP 10 RESTAURANTS LIKE AB's - Absolute Barbecues WITH SIMILAR REVIEWS: 

	cuisines	rating	approx_cost	location
Flechazo	Asian, Mediterranean, North Indian, BBQ	4.9	1400.0	Whitefield
Byg Brewski Brewing Company	Continental, North Indian, Italian, South Indi...	4.9	1600.0	Sarjapur Road
AB's - Absolute Barbecues	European, Mediterranean, North Indian, BBQ	4.8	1600.0	Whitefield
The Black Pearl	North Indian, European, Mediterranean, BBQ	4.8	1500.0	Marathahalli
AB's - Absolute Barbecues	European, Mediterranean, North Indian, BBQ	4.7	1600.0	Sarjapur Road
The Black Pearl	North Indian, European, Mediterranean	4.7	1400.0	Koramangala 5th Block
Hammered	North Indian, Thai, Japanese, Continental, Cafe	4.7	1300.0	Cunningham Road
Buff Buffet Buff	North Indian, Chinese, Continental, Thai, Salad	4.5	1500.0	Koramangala 5th Block
Deja Vu Resto Bar	North Indian, Italian	4.4	900.0	Bannerghatta Road
Barbecoa	North Indian, Continental, Chettinad, Andhra, ...	4.3	1000.0	Marathahalli

we have got a recommendation of top 10 similar restaurants based on their cosine score

Conclusion

In this project, an attempt has been made to understand the restaurant industry in Bengaluru and predict the rating of a restaurant with 98.9% accuracy using a random forest algorithm based on location, availability of online order service, availability of dine-in service, type of cuisines offered, type of service offered, number of votes received, the sentiment of the reviews. Also, a recommender was built based on NLP to suggest similar restaurants for the customers. This will empower restaurants with the data required to take decisions to make their business successful. Future possible research could make use of other significant factors which includes the foot traffic competition i.e the number of similar businesses that could impact the new business being established, accessibility, and average business rates that could be incurred for a particular type of restaurant. These above-mentioned factors could help the system make the analysis more accurate